[Feature Request] Need smaller artifact: Use xz instead gzip to compress them

[jjardon]

This will make compressing them slower, but seems our bottleneck is bandwidth, no cpu power

As a comparison, the same artifact (gcc):

gzip: 152.5 MB
xz (lzma): 78.5 MB

https://docs.python.org/3/library/lzma.html
https://pypi.python.org/pypi/backports.lzma/ (backport for python < 3.3)